OBJECTIVES:

To assess whether preterm infants with postnatal cytomegalovirus infection develop neurologic sequelae in early childhood.

METHODS:

Infants <32 weeks’ gestation were prospectively screened for cytomegalovirus (CMV) at term-equivalent age. Neurodevelopment was compared between CMV-positive and CMV-negative infants by using the Griffiths Mental Development Scales (GMDS) at 16 months’ corrected age (CA); the Bayley Scales of Infant and Toddler Development, Third Edition or the GMDS at 24 to 30 months’ CA; and the Wechsler Preschool and Primary Scale of Intelligence, Third Edition and Movement Assessment Battery for Children, Second Edition at 6 years of age. At 6 years old, hearing was assessed in CMV-positive children.

RESULTS:

Neurodevelopment was assessed in 356 infants at 16 months’ CA, of whom 49 (14%) were infected and 307 (86%) were noninfected. Infected infants performed significantly better on the GMDS locomotor scale. There were no differences at 24 to 30 months’ CA on the Bayley Scales of Infant and Toddler Development, Third Edition or GMDS. At 6 years of age, infected children scored lower on the Wechsler Preschool and Primary Scale of Intelligence, Third Edition, but mean scores were within normal range, reaching significance only in verbal IQ (96 [SD 17] vs 103 [SD 15] points; P = .046). Multiple regression indicated no impact of CMV status but significant influence of maternal education and ethnicity on verbal IQ. No significant differences in motor development were found and none of the infected children developed sensorineural hearing loss.

CONCLUSIONS:

In this cohort study, postnatal cytomegalovirus infection in preterm children did not have an adverse effect on neurodevelopment within the first 6 years of life.

What’s Known on This Subject:

Postnatal cytomegalovirus (pCMV) infection is most frequently transmitted through fresh breast milk. Despite its benefits, breast milk is frequently withheld or pretreated to avoid virus transmission to preterm infants. The effects of a pCMV infection on neurodevelopment are insufficiently studied.

What This Study Adds:

This study does not show an adverse effect of pCMV infection on neurodevelopment, including hearing in infancy and early childhood. Therefore, measures to withhold fresh breast milk in the neonatal period may not be warranted.

Postnatal cytomegalovirus (pCMV) infection is a common viral infection and frequently affects preterm infants (gestational age [GA] <32 weeks) and very low birth weight (VLBW) (birth weight <1500 g) infants, with an estimated median incidence of 20%.1,2 cytomegalovirus (CMV) is mainly transmitted via CMV-seropositive mothers shedding the virus in their breast milk, of whom ∼37% to 76% will transmit the virus to their infants.3,4 Whereas term infants are asymptomatic, preterm infants and VLBW infants may be at risk for symptomatic disease.5,6 Symptomatic disease, such as CMV-related sepsislike syndrome (SLS), thrombocytopenia, pneumonia, and/or hepatitis, may occur but is rare (median incidence 4%).1,5,7 CMV is the leading nongenetic cause of sensorineural hearing loss (SNHL).8 SNHL was not found in infants with pCMV infection at 2 years9 and 8 years of age.10 Data on the neurodevelopmental outcomes of children with pCMV infection are scarce and limited to small-cohort studies. In preterm infants, neurodevelopment until 4 years of age appears to be within the normal range.10,12 However, researchers in several studies have suggested a negative impact on cognitive development at school age.13,15 Because of the uncertainty regarding the short- and long-term consequences of a pCMV infection, fresh, untreated breast milk is not always given to VLBW infants.16,18 Because (untreated) breast milk is known to improve infant health,19 it is important to study the effects of pCMV infection in a large cohort of preterm infants. Our aim in this prospective, longitudinal cohort study is to examine the consequences of a pCMV infection on neurodevelopmental outcomes, including hearing, in a cohort of preterm infants until 6 years of age.

From April 2007 until December 2010, all infants <32 weeks GA admitted to the level 3 NICU of the Wilhelmina Children’s Hospital in Utrecht, the Netherlands, were screened for CMV, predominantly with urine obtained at term-equivalent age (TEA) (40 weeks postconceptional age) during a routine follow-up visit at the outpatient clinic by using CMV polymerase chain reaction (PCR) as previously described.6 Congenital CMV infection was excluded by a negative CMV-PCR result on urine collected within 1 week after birth.6 Exclusion criteria were as follows: absence of urine at TEA, severe cerebral abnormalities (ie, porencephalic cyst, cystic periventricular leukomalacia, posthemorrhagic ventricular dilatation requiring the insertion of a ventricular reservoir or ventriculoperitoneal shunt, or intraventricular hemorrhage [IVH] grade III and IV), chromosomal anomalies, death before TEA, and no parental consent. For the analysis of neurodevelopmental outcomes at 6 years old, only children with a GA of ≤30 weeks were included. The internal review committee of our hospital approved this study.

Clinical and demographic characteristics were collected as previously described.6 Additionally, small for gestational age (SGA) status, socioeconomic status (SES), and maternal education were recorded. SGA was defined as birth weight by GA <10th percentile. Percentiles for our population were obtained from the Dutch perinatal registry.20 SES was determined indirectly by using the SES of the parent’s postal code and was provided by The Netherlands Institute for Social Research.21 A score <−1 was considered low, between −1 and 1 was average, and >1 was high. Symptoms of pCMV disease included SLS, pneumonia, cholestasis, and/or thrombocytopenia.5 Criteria for a diagnosis of symptomatic CMV disease were the presence of SLS, pneumonia, and cholestasis.6 During admission to the NICU and at TEA, cranial ultrasonography was performed as previously described.6 

Neurodevelopmental outcome was routinely assessed by developmental specialists at the outpatient clinic by using the Griffiths Mental Development Scales (GMDS)22 at 16 months’ corrected age (CA) and the Bayley Scales of Infant and Toddler Development, Third Edition (BSITD-III)23 at 24 to 30 months’ CA. The BSITD-III was routinely used in infants <30 weeks GA at 24 to 30 months. When infants had a higher GA (30–32 weeks), the GMDS was used instead. At 6 years of age, all preterm infants ≤30 weeks GA received a routine general pediatric assessment that included motor function using the Movement Assessment Battery for Children, Second Edition (MABC-II).24 Cognitive function was routinely assessed at our hospital by using the Wechsler Preschool and Primary Scale of Intelligence, Third Edition (WPPSI-III)25 for all children born ≤28 weeks GA. Children born >28 weeks GA who have abnormal development at 6 years old are only tested with the WPPSI-III at the request of the pediatrician. Therefore, inclusion numbers are lower for the WPPSI-III than for the MABC-II (Fig 1). All eligible infants with a pCMV infection and >28 weeks GA were tested by using the WPPSI-III as part of this study. A detailed account of the GMDS, BSITD-III, MABC-II, and WPPSI-III has been previously described.26,27 For the GMDS and BSITD-III, z scores were calculated to compare outcome at 2 years of age. Both the GMDS and BSITD-III scores were corrected for preterm birth. When the BSITD-III was administered at 24 months’ CA, parents were asked to provide the age of onset of independent walking (AOIW), which was defined as walking at least 5 steps independently.23 

FIGURE 1

Inclusion of study population. a TEA: term-equivalent age (40 weeks postconceptional age).

FIGURE 1

Inclusion of study population. a TEA: term-equivalent age (40 weeks postconceptional age).

Close modal

At 6 years of age, all children with a pCMV infection underwent pure tone audiometric testing with headphones, following standardized procedures in a tertiary-care audiology center.

When pure tone audiometric testing was not feasible because of patient incompliance, behavioral observation audiometry was used instead. Pure tone–averaged (500, 1000, 2000, and 4000 Hz) hearing loss was calculated for each ear. Audiograms were assessed for conductive or SNHL. Speech performance curves were measured by using the standardized Dutch consonant-vowel-consonant list for children28 and assessed for the maximum speech discrimination score and curve displacement (>20 dB from the reference curve). Middle-ear function was assessed by impedance audiometry. The degree of hearing impairment was classified according to the World Health Organization grading system.29 SNHL was defined as a threshold elevation of >25 dB without any component of conductive hearing loss.

Statistical analysis was performed by using IBM SPSS Statistics version 23 (IBM Corporation). Figures were produced in GraphPad Prism version 5.03 (GraphPad Software Inc, La Jolla, CA). Categorical and dichotomous variables were analyzed by using the χ2 test. Continuous variables were analyzed with 2-tailed Student’s t tests. One-way analysis of variance was used to determine correlations among continuous variables. A P value of <.05 was considered statistically significant. Multiple-regression analyses were performed with AOIW and subsets of the WPPSI-III as dependent variables and pCMV infection and significantly correlated parameters as independent variables.

During the study period, CMV status at TEA could be determined in 462 of 701 (66%) preterm infants <32 weeks GA, of whom 411 (89%) were eligible for follow-up (Fig 1). CMV status could not be determined in 186 of 701 (27%) infants because the collection of urine at TEA was not successful. A pCMV infection was diagnosed in 74 infants; congenital CMV infection was excluded in 63 (85%) by using a CMV-PCR of the urine collected in the first week and in 11 (15%) by using CMV-PCR combined with anti-CMV immunoglobulin-M analysis of dried blood spots cards. Of the 74 infected infants, 56 (76%) could be included in the hearing study. Clinical symptoms of CMV-disease were observed in 4 of 74 (5%) infants and included thrombocytopenia (n = 1), pneumonia (n = 2), and SLS with pneumonia and thrombocytopenia (n = 1). One infant with symptomatic pCMV infection died at 6 months of age because of severe prematurity–related respiratory problems present since birth and before the pCMV infection. None of the infants were treated with Valganciclovir or Ganciclovir because all symptoms were self-limiting.5 All CMV-positive infants were exclusively fed fresh breast milk from their mothers. Clinical and demographic data with respect to CMV status are summarized in Table 1 for infants who were tested at 24 to 30 months and for children who were tested at 6 years of age. The clinical characteristics of infants who were assessed at 16 months were comparable with infants who were assessed at 24 to 30 months (data not shown). Baseline characteristics of the infants who were tested with the WPPSI-III at 6 years of age can be found in Supplemental Table 3. Baseline characteristics of the children who received hearing assessment at 6 years of age can be found in Supplemental Table 4. Overall, infected infants were significantly more often born to mothers of non-Western origin, were fed fresh breast milk more frequently, and had lenticulostriate vasculopathy (LSV) at TEA more frequently (Table 1). These findings have been previously reported.6 

TABLE 1

Clinical and Demographic Characteristics of Preterm Infants Assessed With the BSITD-III or GMDS at 24 Months’ CA and/or 6 Years of Age With the WPPSI-III and/or the MABC-II With Respect to pCMV Infection

Infants Assessed at 24 mo CA (GA <32 wk)PInfants Assessed at 6 y (GA ≤30 wka)P
CMV-Positive (n = 55)CMV-Negative (n = 289)CMV-Positive (n = 41)CMV-Negative (n = 172)
Clinical and demographic characteristics 
 GA, mean, wk, (range) 28.2 (24.3–31.3) 28.8 (24.7–31.9) .022 27.7 (24.3–30) 28 (24.4–30) .242 
 Birth wt, mean, g (SD) 1129 (287) 1181 (319) .261 1076 (266) 1074 (257) .964 
 SGA, n (%) 5 (13) 39 (14) .803 1 (2) 16 (9) .145 
 Male sex, n (%) 31 (56) 157 (54) .781 23 (56) 83 (48) .367 
 NWMO, n (%) 30 (55) 42 (15) <.001 25 (61) 21 (12) <.001 
 Apgar score at 1 min, mean (SD) 7 (2) 7 (2) .231 7 (2) 6 (2) .463 
 Apgar score at 5 min, mean (SD) 8 (2) 8 (1) .592 8 (1) 8 (1) .965 
 Breast milk, n (%) 55 (100) 229 (79) <.001 41 (100) 137 (80) .002 
 RDS, n (%) 22 (40) 150 (52) .106 19 (46) 112 (65) .026 
 >7 d mechanical ventilation, n (%) 3 (6) 45 (16) .047 2 (5) 31 (18) .037 
 Chronic lung disease, n (%) 1 (2) 16 (6) .244 1 (2) 14 (8) .200 
 PDA, n (%) 12 (22) 71 (25) .662 10 (24) 51 (30) .503 
 Surgery for PDA, n (%) 2 (4) 10 (4) .948 2 (5) 7 (4) .817 
 Hypotension treated with inotropes, n (%) 11 (20) 105 (36) .019 6 (15) 72 (42) <.001 
 No. transfusions, median (range) 1 (0–15) 1 (0–18) .881 2 (0–15) 2 (0–18) .598 
 Sepsis, n (%) 13 (24) 94 (33) .192 8 (20) 59 (34) .067 
 Necrotizing enterocolitis, n (%) 9 (3) .185 6 (4) .225 
 NICU admission d, mean (SD) 33 (21) 34 (24) .890 38 (21) 41 (24) .435 
SES 
 Low, n (%) 12 (22) 35 (12) .055 11 (27) 19 (11) .009 
 Average, n (%) 34 (62) 192 (66) .508 23 (56) 116 (67) .170 
 High, n (%) 9 (16) 62 (22) .393 7 (17) 37 (22) .528 
Cranial ultrasonography findings 
IVH (%)       
 Grade I, n (%) 5 (11) 29 (14) .576 5 (12) 20 (12) .919 
 Grade II, n (%) 6 (13) 27 (13) .996 4 (10) 24 (14) .475 
LSV at TEA, n (%) 20 (36) 49 (17) .001 15 (37) 34 (20) .021 
Germinolytic cysts at TEA, n (%) 5 (15) 12 (13) .729 5 (12) 25 (15) .699 
Infants Assessed at 24 mo CA (GA <32 wk)PInfants Assessed at 6 y (GA ≤30 wka)P
CMV-Positive (n = 55)CMV-Negative (n = 289)CMV-Positive (n = 41)CMV-Negative (n = 172)
Clinical and demographic characteristics 
 GA, mean, wk, (range) 28.2 (24.3–31.3) 28.8 (24.7–31.9) .022 27.7 (24.3–30) 28 (24.4–30) .242 
 Birth wt, mean, g (SD) 1129 (287) 1181 (319) .261 1076 (266) 1074 (257) .964 
 SGA, n (%) 5 (13) 39 (14) .803 1 (2) 16 (9) .145 
 Male sex, n (%) 31 (56) 157 (54) .781 23 (56) 83 (48) .367 
 NWMO, n (%) 30 (55) 42 (15) <.001 25 (61) 21 (12) <.001 
 Apgar score at 1 min, mean (SD) 7 (2) 7 (2) .231 7 (2) 6 (2) .463 
 Apgar score at 5 min, mean (SD) 8 (2) 8 (1) .592 8 (1) 8 (1) .965 
 Breast milk, n (%) 55 (100) 229 (79) <.001 41 (100) 137 (80) .002 
 RDS, n (%) 22 (40) 150 (52) .106 19 (46) 112 (65) .026 
 >7 d mechanical ventilation, n (%) 3 (6) 45 (16) .047 2 (5) 31 (18) .037 
 Chronic lung disease, n (%) 1 (2) 16 (6) .244 1 (2) 14 (8) .200 
 PDA, n (%) 12 (22) 71 (25) .662 10 (24) 51 (30) .503 
 Surgery for PDA, n (%) 2 (4) 10 (4) .948 2 (5) 7 (4) .817 
 Hypotension treated with inotropes, n (%) 11 (20) 105 (36) .019 6 (15) 72 (42) <.001 
 No. transfusions, median (range) 1 (0–15) 1 (0–18) .881 2 (0–15) 2 (0–18) .598 
 Sepsis, n (%) 13 (24) 94 (33) .192 8 (20) 59 (34) .067 
 Necrotizing enterocolitis, n (%) 9 (3) .185 6 (4) .225 
 NICU admission d, mean (SD) 33 (21) 34 (24) .890 38 (21) 41 (24) .435 
SES 
 Low, n (%) 12 (22) 35 (12) .055 11 (27) 19 (11) .009 
 Average, n (%) 34 (62) 192 (66) .508 23 (56) 116 (67) .170 
 High, n (%) 9 (16) 62 (22) .393 7 (17) 37 (22) .528 
Cranial ultrasonography findings 
IVH (%)       
 Grade I, n (%) 5 (11) 29 (14) .576 5 (12) 20 (12) .919 
 Grade II, n (%) 6 (13) 27 (13) .996 4 (10) 24 (14) .475 
LSV at TEA, n (%) 20 (36) 49 (17) .001 15 (37) 34 (20) .021 
Germinolytic cysts at TEA, n (%) 5 (15) 12 (13) .729 5 (12) 25 (15) .699 

PDA, patent ductus arteriosus.

a

GA of 30 completed weeks or less.

A total of 49 of 356 (14%) infants with a pCMV infection and 307 of 356 (86%) noninfected infants were tested at a mean age of 16.4 months’ CA (SD 1.8) and 16.1 months’ CA (SD 1.3), respectively. The mean locomotor subscale quotient was significantly higher in infants with a pCMV infection (z score 0.35 [SD 0.81] vs 0.02 [SD 0.98] in the control group, which corresponds to a quotient of 102 [SD 12] vs 97 [SD 14]; P = .025; Fig 2A). Other subscales did not differ significantly. The mean general developmental quotient was comparable in infected and noninfected infants (102 [SD 9] vs 101 [SD 9], respectively; P = .320) and was within the normal range.

FIGURE 2

A, z score of the GMDS at 16 months’ CA. B, z score of the BSITD-III at 24 months’ CA. C, z score of the GMDS at 30 months’ CA. The upper and lower borders of the box plots represent the 25th and 75th percentiles; bars represent median z scores, and whiskers represent the 5th and 95th percentiles. Outliers are depicted as dots.

FIGURE 2

A, z score of the GMDS at 16 months’ CA. B, z score of the BSITD-III at 24 months’ CA. C, z score of the GMDS at 30 months’ CA. The upper and lower borders of the box plots represent the 25th and 75th percentiles; bars represent median z scores, and whiskers represent the 5th and 95th percentiles. Outliers are depicted as dots.

Close modal

A total of 49 of 289 (17%) infants with a pCMV infection and 240 of 289 (83%) noninfected infants were tested by using the BSITD-III at a mean age of 26.1 months’ CA (SD 3.0) and 25.3 months’ CA (SD 2.5), respectively. In 6 of 55 (11%) infected infants and 49 of 55 (89%) noninfected infants, the GMDS was used instead at a mean age of 23.1 months (SD 1.2) and 24.6 months (SD 1.7), respectively. There were no significant differences in BSITD-III and GMDS z scores between infected and noninfected infants (Fig 2B and 2C, respectively). The mean corrected cognitive composite scores and total motor composite score in infected and noninfected infants by using the BSITD-III at 24 months (104 [SD 10] vs 105 [SD 12], respectively; P = .184; 109 [SD 10] vs 109 [SD 12], respectively; P = .748) were within the normal range.

The mean AOIW was compared between 49 infants with a pCMV infection and 239 noninfected infants. Postnatally infected infants were able to walk at a younger CA compared with noninfected infants (14.7 months [SD 2.4] and 15.8 months [SD 3.1], respectively; P = .026). Multiple regression analysis showed that significantly earlier AOIW was related to non-Western maternal origin (NWMO) (Supplemental Table 5). The mean AOIW was 14.1 months (SD 2.8) in infants of mothers of non-Western ethnicity compared with a mean AOIW of 16.0 months (SD 3.0) in infants of mothers of Western ethnicity (P < .001).

In total, 33 of 118 (28%) children with a pCMV infection and 85 of 118 (72%) noninfected children were tested at a mean age of 5.7 years (SD 0.3) and 5.7 years (SD 0.5), respectively. Mean scores were in the normal range on all 4 domains for infected and noninfected children, but infected children had overall lower scores (Table 2). This only reached statistical significance on the subscale of verbal IQ (96 [95% confidence interval (CI) 90 to 102] vs 103 [95% CI 100 to 106]; P = .046). Multiple-regression analyses indicated that the presence of respiratory distress syndrome (RDS) significantly impacted total IQ (coefficient −5.9; P = .046) and verbal IQ (coefficient −8.8; P = .004; Supplemental Table 6). NWMO (coefficient −7.7; P = .049) and low maternal education (coefficient −8.3; P = .019) also significantly impacted verbal IQ. CMV was not significantly associated with impaired outcomes on any of the 4 WPPSI-III domains. No significant differences were found between symptomatic and asymptomatic children (Supplemental Tables 7 and 8).

TABLE 2

Results of the WPPSI-III and MABC-II at 6 Years of Age

MeasureP
WPPSI-III CMV-positive (n = 33) CMV-negative (n = 85)  
 Total IQ, mean (95% CI)a 96 (91 to 101) 101 (98 to 104) .079 
 Verbal IQ, mean (95% CI) 96 (90 to 102) 103 (100 to 106) .046 
 Performance IQ, mean (95% CI) 100 (95 to 105) 102 (99 to 105) .456 
 Processing speed quotient, mean (95% CI)b 92 (86 to 98) 93 (90 to 96) .708 
 
MABC-II CMV-positive (n = 39) CMV-negative (n = 167)  
 Total impairment score, median (range)a 8 (4–14) 9 (2–15) .661 
 Manual dexterity, median (range) 8 (4–14) 9 (2–15) .902 
 Aiming and catching ball skills, median (range) 10 (5–17) 9 (1–15) .052 
 Dynamic and static balance, median (range)a 8 (4–16) 9 (2–17) .081 
MeasureP
WPPSI-III CMV-positive (n = 33) CMV-negative (n = 85)  
 Total IQ, mean (95% CI)a 96 (91 to 101) 101 (98 to 104) .079 
 Verbal IQ, mean (95% CI) 96 (90 to 102) 103 (100 to 106) .046 
 Performance IQ, mean (95% CI) 100 (95 to 105) 102 (99 to 105) .456 
 Processing speed quotient, mean (95% CI)b 92 (86 to 98) 93 (90 to 96) .708 
 
MABC-II CMV-positive (n = 39) CMV-negative (n = 167)  
 Total impairment score, median (range)a 8 (4–14) 9 (2–15) .661 
 Manual dexterity, median (range) 8 (4–14) 9 (2–15) .902 
 Aiming and catching ball skills, median (range) 10 (5–17) 9 (1–15) .052 
 Dynamic and static balance, median (range)a 8 (4–16) 9 (2–17) .081 
a

Data from 1 child missing.

b

Data from 6 children missing.

Thirty-nine (19%) children with a pCMV infection and 167 (81%) noninfected children were tested at a mean age of 5.8 years (SD 0.2) and 5.7 years (SD 0.4), respectively (P = .155). There were no significant differences in the median scores of the MABC-II subscales between both groups (Table 2). The median total impairment standard score between infected children and noninfected children was 8 points and 9 points (P = .661), respectively, and was in the normal range.

Seventy-four (16%) children had a pCMV infection, of whom 56 (76%) were audiologically tested at a mean age of 5.8 years (SD 0.2; Fig 1). None of the children had SNHL. Pure tone audiometry was conducted in all children. Forty-four (79%) children had normal hearing. A slight hearing impairment (26–40 dB) was seen in 9 (16%) children, of whom 7 had unilateral conductive hearing loss and 2 had bilateral conductive hearing loss. Tympanometry indicated middle-ear dysfunction in all of them (middle-ear fluid, n = 2; negative pressure in the middle-ear cavity, n = 5; middle-ear fluid and negative pressure in the middle-ear cavity, n = 2). A moderate impairment (41–60 dB) was seen in 3 (5%) children, of whom 1 had unilateral conductive hearing loss and 2 had bilateral conductive hearing loss. Again, tympanometry indicated middle-ear dysfunction in all children (middle-ear fluid, n = 1; negative pressure in the middle-ear space, n = 1; cholesteatoma and negative pressure in the middle-ear space, n = 1). Speech audiometry was conducted in 2 children and reflected the conductive moderate impairment seen on pure tone audiometry. Cochlear function was normal in all the tested children.

To the best of our knowledge, this is the largest prospective cohort study in which researchers examine neurodevelopmental outcomes of preterm infants with a pCMV infection from birth until early childhood. The results of this study did not show an impaired neurodevelopmental outcome (including SNHL) until 6 years of age in preterm infants with a pCMV infection. So far, outcome studies were inconclusive in terms of the long-term effects of a pCMV infection in preterm infants. Although short-term development until 2 to 4.5 years of age seems to be unaffected,11,12,30 more recently, smaller case-controlled studies managing infected children until school age and into adolescence have shown subtle impairments in cognitive functioning when compared with uninfected controls.10,13,14,31 In line with previous studies on short-term outcomes,11,12,30 we did not find any adverse neurodevelopmental sequelae in 55 preterm infants with a pCMV infection compared with 289 noninfected infants until 2 years’ CA. Cognitive assessment at 6 years of age indicated slight discrepancies between the groups. Overall, infected children scored in the normal range but had lower mean scores on all domains, reaching significance on the domain of verbal IQ only. CMV status did not significantly contribute to the observed variance on any subscales of the WPPSI-III. NWMO and low maternal education were significantly associated with a lower verbal IQ score. NWMO has been previously identified as a risk factor for pCMV acquisition,6 and children with pCMV infection in this study had significantly more mothers of non-Western ethnicity.5 It is conceivable that children whose parents do not speak Dutch as their first language may learn the native languages and differing cultural norms of their parents’ ethnic backgrounds first.32,33 As such, the WPPSI-III, which is conducted in Dutch at our institution, may pose a cultural and/or language barrier to these children. The occurrence of pCMV infection and low verbal IQ may not be causal but rather unrelated occurrences associated with social and epidemiologic factors. Several long-term, developmental outcome studies of children with a pCMV infection reported similar cognitive results at 6 and 8 years of age.10,13 In contrast to our results, however, in the study by Bevot et al,10 pCMV infection contributed independently to lower scores on all cognitive subscales and, in combination with paternal SES, on the overall cognitive composite score at 8 years of age. Interestingly, in the study by Goelz et al,13 which included the cohort of Bevot et al,10 the significant cognitive differences initially observed were not confirmed. Similar to our study, low SES (consisting also of parental education) significantly contributed to the overall lower cognitive scores. The absolute contribution of CMV status on outcomes was not reported. It has been suggested that the effects of an early pCMV infection may only manifest at school age and/or adolescence, a time when complex, higher cognitive functions develop.14 In the most recent study by Brecht et al,14 infected adolescents (11–17 years of age) had significantly lower general intelligence scores than noninfected preterm and term adolescents, findings that could not be explained by differences in maternal education, attention, or brain pathology. An important confounder not controlled for in the study by Brecht et al14 is parental ethnicity. Lower maternal education and an immigrant background of both parents have been linked to an impaired long-term composite IQ at 10 to 13 years of age.34 

Researchers in a recent study observed an increased occurrence of bronchopulmonary dysplasia in VLBW infants with a pCMV infection.35 In the current study, the development of bronchopulmonary dysplasia was similar in both groups, in contrast to the findings by Kelly et al,35 most likely because of methodological differences in study design. Selection bias may have confounded the results of Kelly et al35 because they were not weighed against a population of asymptomatic VLBW infants with a pCMV infection; therefore, these results should be interpreted with caution. In our population studied at 6 years of age, the control group had significantly more RDS and mechanical ventilation for >7 days. RDS independently contributed to a lower total IQ and verbal IQ, but still the control group attained a higher score compared with infected infants. It is possible that the differences in scores between the groups would have been more pronounced without baseline differences in neonatal morbidity (including RDS).

In terms of motor development, infected infants had better gross motor performance at 16 months’ CA and had significantly earlier AOIW. We analyzed the association between pCMV infection and AOIW for possible confounders and found that NWMO ethnicity had a positive effect on independent walking. Ethnicity has previously been associated with AOIW in a study on AOIW in Dutch preterm infants.36 At 6 years of age, all median scores were in the normal range, without significant differences on all subscales of the MABC-II. We have previously demonstrated microstructural changes in the occipital white matter of infants with a pCMV infection.37 Although no visual assessment was conducted, infected children actually scored better than noninfected children on ball skills when using the MABC-II at 6 years of age. The significance of these microstructural changes remains to be elucidated. Previously and in the current study, we have shown that LSV not yet present at birth is more common in preterm infants with a pCMV infection.6 LSV is not pathognomonic for CMV infections, and the exact causal relationship between pCMV infection and the occurrence of LSV remains unclear.38 Previous studies in infants with a pCMV infection have reported an association between pCMV infection and the presence of LSV.39,40 In this study, LSV was not associated with impaired neurodevelopmental outcome at 2 years’ CA and at 6 years of age.

SNHL is a well-known sequela of pCMV infection8 but has so far not been detected in infants with a pCMV infection.5 At 12 and 24 months, SNHL has been previously excluded for this cohort.9 At 6 years of age, despite the presence of conductive hearing loss among 12 children, none of the infected children developed SNHL.12 

This study has several limitations, most important of them being that we did not determine the onset of CMV infection, and therefore, we cannot exclude that an early pCMV infection may still have detrimental effects on neurodevelopment. The timing of CMV acquisition and subsequent first viral detection in relation to birth weight and GA seem to be the most important risk factors for symptomatic CMV disease.7,41,42 In this study, of the infants with a pCMV infection assessed at 6 years of age (n = 33), 9% (n = 3) were symptomatic in the neonatal period. No significant difference was noted between symptomatic and asymptomatic infants (Supplemental Tables 7 and 8). Because so few infants were symptomatic, no conclusions can be made about the impact of clinical symptoms on outcomes. In the subgroup of noninfected infants >28 weeks GA, selection bias may have been introduced because the WPPSI-III was performed at the discretion of the pediatrician. This selection may have had a negative effect on the WPPSI-III scores. Nevertheless, WPPSI-III scores in this subgroup were within the normal range of the population (data not shown). Furthermore, there was the common issue of loss to follow-up in long-term, prospective follow-up studies, which may have introduced selection bias. More prospective cohort studies are needed whereby the timing of virus acquisition and symptomatic disease are correlated with long-term neurodevelopmental outcomes. Despite the unsuccessful collection of urine among 27% of the infants during visits to the outpatient clinic at TEA, a comparative analysis at baseline did not show any statistically significant differences compared with the included infants.

The results of the current study, in which no obvious deleterious effect of pCMV infection on neurodevelopmental outcomes in early childhood has been shown, may not justify interventions like pasteurization or freezing or withholding breast milk to prevent CMV transmission in the general preterm population.

Different feeding policies have been observed in NICUs among several European countries whereby mothers are often routinely tested for CMV serostatus, and, if positive, the breast milk is pretreated, formula is given,18 or routine, standard freezing of all breast milk for infants <32 weeks GA is performed.16 In extremely preterm and VLBW infants, however, CMV may act as an aggravator of an already fragile system, causing symptomatic infection.5,7 These infants may benefit from a delayed introduction of breast milk or pretreatment of breast milk to prevent a symptomatic pCMV infection5 until the exact long-term consequences are determined.

The neurodevelopmental outcomes of children with a pCMV infection was within the normal range in early childhood. At 6 years of age, infected children had lower cognitive scores, with only a significant difference in verbal IQ, which could be attributed to maternal education and ethnicity and not CMV status. Median motor function at 6 years of age was within the normal range. None of the infected children developed perceptive hearing loss. More prospective cohort studies are needed to examine cognitive development in extremely preterm infants (GA <26 weeks) to determine the consequences of early and/or symptomatic pCMV infection.

     
  • AOIW

    age of onset of independent walking

  •  
  • BSITD-III

    Bayley Scales of Infant and Toddler Development, Third Edition

  •  
  • CA

    corrected age

  •  
  • CI

    confidence interval

  •  
  • CMV

    cytomegalovirus

  •  
  • GA

    gestational age

  •  
  • GMDS

    Griffiths Mental Development Scales

  •  
  • IVH

    intraventricular hemorrhage

  •  
  • LSV

    lenticulostriate vasculopathy

  •  
  • MABC-II

    Movement Assessment Battery for Children, Second Edition

  •  
  • NWMO

    non-Western maternal origin

  •  
  • pCMV

    postnatal cytomegalovirus

  •  
  • PCR

    polymerase chain reaction

  •  
  • RDS

    respiratory distress syndrome

  •  
  • SES

    socioeconomic status

  •  
  • SGA

    small for gestational age

  •  
  • SLS

    sepsislike syndrome

  •  
  • SNHL

    sensorineural hearing loss

  •  
  • TEA

    term-equivalent age

  •  
  • VLBW

    very low birth weight

  •  
  • WPPSI-III

    Wechsler Preschool and Primary Scale of Intelligence, Third Edition

Dr Gunkel was responsible for data collection and analysis, drafted the manuscript, and reviewed and finalized the manuscript; Drs de Vries, Jongmans, Koopman-Esseboom, van Haastert, Eijsermans, and van Stam conducted part of the neurodevelopmental testing and critically reviewed the manuscript; Dr van Zanten conducted the audiological testing at 6 years of age and reviewed the manuscript; Dr Wolfs critically reviewed the manuscript; Dr Nijman was responsible for data collection and analysis, manuscript drafting, and critically reviewed the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

FUNDING: Funded by the University Medical Center Utrecht in the Netherlands and the Dutch Phelps Foundation.

We acknowledge and thank Dr M. A. Maciolek-Verboon for her expertise in conceptualizing and conducting this study as well as critically reviewing the results (at 16 months and 24–30 months) and the article.

1
Kurath
S
,
Halwachs-Baumann
G
,
Müller
W
,
Resch
B
.
Transmission of cytomegalovirus via breast milk to the prematurely born infant: a systematic review.
Clin Microbiol Infect
.
2010
;
16
(
8
):
1172
1178
[PubMed]
2
Luck
S
,
Sharland
M
.
Postnatal cytomegalovirus: innocent bystander or hidden problem?
Arch Dis Child Fetal Neonatal Ed
.
2009
;
94
(
1
):
F58
F64
[PubMed]
3
Hamprecht
K
,
Maschmann
J
,
Vochem
M
,
Dietz
K
,
Speer
CP
,
Jahn
G
.
Epidemiology of transmission of cytomegalovirus from mother to preterm infant by breastfeeding.
Lancet
.
2001
;
357
(
9255
):
513
518
[PubMed]
4
Peckham
CS
,
Johnson
C
,
Ades
A
,
Pearl
K
,
Chin
KS
.
Early acquisition of cytomegalovirus infection.
Arch Dis Child
.
1987
;
62
(
8
):
780
785
[PubMed]
5
Gunkel
J
,
Wolfs
TF
,
de Vries
LS
,
Nijman
J
.
Predictors of severity for postnatal cytomegalovirus infection in preterm infants and implications for treatment.
Expert Rev Anti Infect Ther
.
2014
;
12
(
11
):
1345
1355
[PubMed]
6
Nijman
J
,
de Vries
LS
,
Koopman-Esseboom
C
,
Uiterwaal
CS
,
van Loon
AM
,
Verboon-Maciolek
MA
.
Postnatally acquired cytomegalovirus infection in preterm infants: a prospective study on risk factors and cranial ultrasound findings.
Arch Dis Child Fetal Neonatal Ed
.
2012
;
97
(
4
):
F259
F263
[PubMed]
7
Mehler
K
,
Oberthuer
A
,
Lang-Roth
R
,
Kribs
A
.
High rate of symptomatic cytomegalovirus infection in extremely low gestational age preterm infants of 22-24 weeks’ gestation after transmission via breast milk.
Neonatology
.
2014
;
105
(
1
):
27
32
[PubMed]
8
Nance
WE
,
Lim
BG
,
Dodson
KM
.
Importance of congenital cytomegalovirus infections as a cause for pre-lingual hearing loss.
J Clin Virol
.
2006
;
35
(
2
):
221
225
[PubMed]
9
Nijman
J
,
van Zanten
BG
,
de Waard
A-KM
,
Koopman-Esseboom
C
,
de Vries
LS
,
Verboon-Maciolek
MA
.
Hearing in preterm infants with postnatally acquired cytomegalovirus infection.
Pediatr Infect Dis J
.
2012
;
31
(
10
):
1082
1084
[PubMed]
10
Bevot
A
,
Hamprecht
K
,
Krägeloh-Mann
I
,
Brosch
S
,
Goelz
R
,
Vollmer
B
.
Long-term outcome in preterm children with human cytomegalovirus infection transmitted via breast milk.
Acta Paediatr
.
2012
;
101
(
4
):
e167
e172
[PubMed]
11
Jim
W-T
,
Chiu
N-C
,
Ho
C-S
, et al
.
Outcome of preterm infants with postnatal cytomegalovirus infection via breast milk: a two-year prospective follow-up study.
Medicine (Baltimore)
.
2015
;
94
(
43
):
e1835
[PubMed]
12
Vollmer
B
,
Seibold-Weiger
K
,
Schmitz-Salue
C
, et al
.
Postnatally acquired cytomegalovirus infection via breast milk: effects on hearing and development in preterm infants.
Pediatr Infect Dis J
.
2004
;
23
(
4
):
322
327
[PubMed]
13
Goelz
R
,
Meisner
C
,
Bevot
A
,
Hamprecht
K
,
Kraegeloh-Mann
I
,
Poets
CF
.
Long-term cognitive and neurological outcome of preterm infants with postnatally acquired CMV infection through breast milk.
Arch Dis Child Fetal Neonatal Ed
.
2013
;
98
(
5
):
F430
F433
[PubMed]
14
Brecht
KF
,
Goelz
R
,
Bevot
A
,
Krägeloh-Mann
I
,
Wilke
M
,
Lidzba
K
.
Postnatal human cytomegalovirus infection in preterm infants has long-term neuropsychological sequelae.
J Pediatr
.
2015
;
166
(
4
):
834.e1
839.e1
[PubMed]
15
Paryani
SG
,
Yeager
AS
,
Hosford-Dunn
H
, et al
.
Sequelae of acquired cytomegalovirus infection in premature and sick term infants.
J Pediatr
.
1985
;
107
(
3
):
451
456
[PubMed]
16
Omarsdottir
S
,
Casper
C
,
Akerman
A
,
Polberger
S
,
Vanpée
M
.
Breastmilk handling routines for preterm infants in Sweden: a national cross-sectional study.
Breastfeed Med
.
2008
;
3
(
3
):
165
170
[PubMed]
17
Zwiauer
K
,
Deutsch
J
,
Goriup
U
, et al
.
Prävention von muttermilchmediierten CMV-infektionen bei frühgeborenen.
Monatsschr Kinderheilkd
.
2003
;
151
(
12
):
1346
1347
18
Buxmann
H
,
Falk
M
,
Goelz
R
,
Hamprecht
K
,
Poets
CF
,
Schloesser
RL
.
Feeding of very low birth weight infants born to HCMV-seropositive mothers in Germany, Austria and Switzerland.
Acta Paediatr
.
2010
;
99
(
12
):
1819
1823
[PubMed]
19
Johnston
M
,
Landers
S
,
Noble
L
,
Szucs
K
,
Viehmann
L
;
Section on Breastfeeding
.
Breastfeeding and the use of human milk.
Pediatrics
.
2012
;
129
(
3
). Available at: www.pediatrics.org/cgi/content/full/129/3/e827
[PubMed]
20
Visser
GH
,
Eilers
PH
,
Elferink-Stinkens
PM
,
Merkus
HM
,
Wit
JM
.
New Dutch reference curves for birthweight by gestational age.
Early Hum Dev
.
2009
;
85
(
12
):
737
744
[PubMed]
21
Sociaal en Cultureel Planbureau (SCP)
. Rangorde naar sociale status van postcodegebieden in Nederland.
2010
. Available at: https://www.scp.nl/Onderzoek/Lopend_onderzoek/A_Z_alle_lopende_onderzoeken/Statusscores
22
Griffiths
R
.
The Abilities of Young Children: A Comprehensive System of Mental Measurement for the First Eight Years of Life
. Revised ed.
Oxford, United Kingdom
:
Test Agency Ltd
;
1984
23
Bayley
N
.
Bayley Scales of Infant and Toddler Development
. 3rd ed.
San Antonio, TX
:
Harcourt Assessment, Inc
;
2006
24
Henderson
S
,
Sugden
D
,
Bartnett
A
.
Movement Assessment Battery for Children (Movement ABC-2), Examiner’s Manual
. 2nd ed.
London, United Kingdom
:
Harcourt Assessment
;
2007
25
Wechsler
D
.
WPPSI-III-NL Nederlandse Bewerking: Technische Handleiding
. 3rd ed.
Amsterdam, The Netherlands
:
Pearson Assessment and Information BV
;
2010
26
Van Haastert
IC
,
Groenendaal
F
,
Van de Waarsenburg
MK
, et al
.
Active head lifting from supine in early infancy: an indicator for non-optimal cognitive outcome in late infancy.
Dev Med Child Neurol
.
2012
;
54
(
6
):
538
543
[PubMed]
27
Keunen
K
,
Išgum
I
,
van Kooij
BJM
, et al
.
Brain volumes at term-equivalent age in preterm infants: imaging biomarkers for neurodevelopmental outcome through early school age.
J Pediatr
.
2016
;
172
:
88
95
[PubMed]
28
Bosman
AJ
,
Smoorenburg
GF
.
Intelligibility of Dutch CVC syllables and sentences for listeners with normal hearing and with three types of hearing impairment.
Audiology
.
1995
;
34
(
5
):
260
284
[PubMed]
29
World Health Organization
. Grades of hearing impairment. Available at: www.who.int/pbd/deafness/hearing_impairment_grades/en/. Accessed January 17, 2017
30
Miron
D
,
Brosilow
S
,
Felszer
K
, et al
.
Incidence and clinical manifestations of breast milk-acquired Cytomegalovirus infection in low birth weight infants.
J Perinatol
.
2005
;
25
(
5
):
299
303
[PubMed]
31
Dorn
M
,
Lidzba
K
,
Bevot
A
,
Goelz
R
,
Hauser
T-K
,
Wilke
M
.
Long-term neurobiological consequences of early postnatal hCMV-infection in former preterms: a functional MRI study.
Hum Brain Mapp
.
2014
;
35
(
6
):
2594
2606
[PubMed]
32
Prevoo
MJ
,
Malda
M
,
Mesman
J
, et al
.
Predicting ethnic minority children’s vocabulary from socioeconomic status, maternal language and home reading input: different pathways for host and ethnic language.
J Child Lang
.
2014
;
41
(
5
):
963
984
[PubMed]
33
Mistry
RS
,
Biesanz
JC
,
Chien
N
,
Howes
C
,
Benner
AD
.
Socioeconomic status, parental investments, and the cognitive and behavioral outcomes of low-income children from immigrant and native households.
Early Child Res Q
.
2008
;
23
(
2
):
193
212
34
Voss
W
,
Jungmann
T
,
Wachtendorf
M
,
Neubauer
AP
.
Long-term cognitive outcomes of extremely low-birth-weight infants: the influence of the maternal educational background.
Acta Paediatr
.
2012
;
101
(
6
):
569
573
[PubMed]
35
Kelly
MS
,
Benjamin
DK
,
Puopolo
KM
, et al
.
Postnatal cytomegalovirus infection and the risk for bronchopulmonary dysplasia.
JAMA Pediatr
.
2015
;
169
(
12
):
e153785
[PubMed]
36
Nuysink
J
,
van Haastert
IC
,
Eijsermans
MJC
, et al
.
Prediction of gross motor development and independent walking in infants born very preterm using the Test of Infant Motor Performance and the Alberta Infant Motor Scale.
Early Hum Dev
.
2013
;
89
(
9
):
693
697
[PubMed]
37
Nijman
J
,
Gunkel
J
,
de Vries
LS
, et al
.
Reduced occipital fractional anisotropy on cerebral diffusion tensor imaging in preterm infants with postnatally acquired cytomegalovirus infection.
Neonatology
.
2013
;
104
(
2
):
143
150
[PubMed]
38
Leijser
LM
,
Steggerda
SJ
,
de Bruïne
FT
, et al
.
Lenticulostriate vasculopathy in very preterm infants.
Arch Dis Child Fetal Neonatal Ed
.
2010
;
95
(
1
):
F42
F46
[PubMed]
39
Hong
SY
,
Yang
JJ
,
Li
SY
,
Lee
IC
.
Lenticulostriate vasculopathy in brain ultrasonography is associated with cytomegalovirus infection in newborns.
Pediatr Neonatol
.
2015
;
56
(
6
):
408
414
[PubMed]
40
Giannattasio
A
,
Di Costanzo
P
,
Milite
P
, et al
.
Is lenticulostriated vasculopathy an unfavorable prognostic finding in infants with congenital cytomegalovirus infection?
J Clin Virol
.
2017
;
91
:
31
35
[PubMed]
41
Maschmann
J
,
Hamprecht
K
,
Dietz
K
,
Jahn
G
,
Speer
CP
.
Cytomegalovirus infection of extremely low-birth weight infants via breast milk.
Clin Infect Dis
.
2001
;
33
(
12
):
1998
2003
[PubMed]
42
Lanzieri
TM
,
Dollard
SC
,
Josephson
CD
,
Schmid
DS
,
Bialek
SR
.
Breast milk-acquired cytomegalovirus infection and disease in VLBW and premature infants.
Pediatrics
.
2013
;
131
(
6
). Available at: www.pediatrics.org/cgi/content/full/131/6/e1937
[PubMed]

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: Dr Gunkel was financially supported by the Dutch Phelps Foundation (grant 2012–020).

Supplementary data